Welcome to the MADUV Challenge 2025!
The MADUV Challenge is a pioneering initiative for INTERSPEECH 2025. This challenge invites participants to leverage cutting-edge speech and signal processing and machine learning techniques to classify mice as either wild-type or modelling ASD (characteristics), advancing both computer science and biomedical research.
Objective and Motivation
By addressing this classification problem by using mice vocalisations, participants can contribute to developing innovative models that benefit the research on animal models of autistic characteristics and the potential of similarities found in human speech.
Besides, understanding ultrasonic vocal patterns in mice offers a unique perspective on animal neuropsychiatric disorder research. Animal models are valuable in the study of the underlying causes of various characteristics and behaviours in conditions like autism, and the ability to classify behaviours from vocal patterns could pave the way for new diagnostic tools and support strategies.
Significance of Participation
Gaining recognition in the INTERSPEECH community.
Advancing machine learning techniques in non-traditional domains.
Contributing to research that may lead to a better understanding of animal models of autistic characteristics.
Organising Team
The MADUV Challenge is brought to you by an esteemed team of researchers from The University of Tokyo, Technical University of Munich, and Kobe University, all renowned for their contributions to speech technology, biomedical research, and affective computing. All organisers are introduced here.
(c) 2025 - Educational Physiology Laboratory,
Graduate School of Education, The University of Tokyo